i'm on a roll... the kgsshv

[hhobeika]

3 hours ago, Laowei said:

One source I used. Great to deal with.

https://www.ebay.com/itm/PTFE-wire-Alpha-5854-7-or-equiv-20-ga-Mil-16878-4-silver-plated-10-feet/292416116301

Many other wire gauges, solid or stranded available by seller. Do a search of his other items.

I second this ebay seller as I initially purchased from him.  He has large variety of colors.

The hook-up wire I used is multi-stranded PTFE #22 for all but the earth where I used #18.

Another source that sells per foot that I also used is Remington Industries either directly through their website or through amazon (www.remingtonindustries.com).

There is also Bulk Wire (bulkwire.com) but I haven't purchased from them.

 

[hhobeika]

Since photography is a hobby of mine as well, I decided to take some internal and external pictures of my recently completed KGSSHV.  Hope you enjoy them.

 

Please click or copy/paste the Dropbox link below to view those pictures:

https://www.dropbox.com/sh/4ox0xh7s1aiw5cb/AACdu5DpeRyeFuINloC8HkRFa?dl=0

[insanity]

Wow what a beauty! Very nicely done, internally as well as the chassis.

Congrats!

[eslover]

It took me 7 years to build my second kgsshv, or so I thought, and no cigar as I powered this one up. I have a problem with the positive, HV rail in the power supply: I measured ~10V instead of +450V. I need some help to troubleshoot.

Couple of things:

1) when powering on, after long discharge time, I get sparks in the neighborhood of the 680uF/450V caps for that rail. This lasts for a second, and goes away. First time i switched off right away, but when I turned amp back on soon after, no problem... if one waits more than 30minutes, then small sparks come back on first “ignition”. No smoke.
1b) After close visual inspection, two legs of the bridge diode for this side of the PS are a bit darker than the others (bronze close to pcb rather than silver). These are the legs connected to the top plane of the pcb. Using a multimeter, diodes seem to be ok for allowing current flow in one direction and not the other (using diode checker on multimeter). There seemed to be missing copper pads top of board (if there was any). Above the 0 in the rev0.9.

1c) I confirmed the two legs, one per diode, are not connected to top plane. Did I vaporize copper at first power up, or was there even any (I assembled the ps years ago, never checked before assembly as I solder from bottom face)

2) output of transformer measures 493V AC. (Both sides)

3) negative rail of PS OK (-453V out of PS).

4) PS board is regular size (not mini), rev v0.9.

Any help would be greatly appreciated. Thanks.

Edited December 13, 2020 by eslover
Added details 1b and 1c

[Rodeodave]

Can you post a pic of the whole PSU so we can see cap and diode orientation and voltage references and so on?

[Pars]

Isn't the use of 450V caps pushing it with respect to your voltages? Your inputs are 493Vac, and your output is 453Vdc? I would think that 550V caps would be mandatory for your use case.

[spritzer]

The caps are in series.  You can even use 350V caps in most cases... 

[Pars]

Sorry, was thinking GRHV...

[Pars]

On 11/1/2020 at 10:50 PM, hhobeika said:

Here is a picture of the inside.  There is a limit on the file size I can upload so I had to dramatically shrink it and compress it to be able to post it.

I had looked at this closer awhile back, and noticed what looks like a wire loom or separator down on the floor of the amp by the output jacks, to hold the output lines in order. Is this something that you made, or is commercially available? I did a quick search at Mouser but didn't turn up anything like this. Thanks!

[hhobeika]

22 hours ago, Pars said:

I had looked at this closer awhile back, and noticed what looks like a wire loom or separator down on the floor of the amp by the output jacks, to hold the output lines in order. Is this something that you made, or is commercially available? I did a quick search at Mouser but didn't turn up anything like this. Thanks!

Those are Nylon Caterpillar Grommet Strips repurposed as wire spacers.  I cut them to the number of slots needed.  Each is held in place using a tie wrap and a wire tie mount.  You can find them at Mouser by searching under "grommet strip".  I had some lying around so I can't give a specific part number but they come in various sizes.

 

I only used them on the output wires to keep them spaced in order to minimize the capacitance.

More photos of my build at the Dropbox link below:

https://www.dropbox.com/sh/4ox0xh7s1aiw5cb/AACdu5DpeRyeFuINloC8HkRFa?dl=0

Hope this helps.

[Pars]

Thanks for replying. Very clever use of the grommet strips. Once you brought them up, I recall seeing these used for grommet holes. I would have never thought of using them for this purpose. Well, maybe if I had some on hand, a light bulb would have gone on...

[powertoold]

My 500V offboard kgsshv currently has two rev 0.4 amp boards and a rev 0.6 power board.

I am looking to swap one of the rev 0.4 amp boards to a rev 0.5 board.

When I change from a rev 0.4 to a rev 0.5 amp board, I will need to:

1) Not cross the 50K resistors

2) Not cross the 1n914 diodes

3) Not flip the O+ and O- bias outputs

Is this correct and also is there anything else I need to do when going from rev 0.4 to rev 0.5 boards?

Thanks y'all!

Edited January 21, 2021 by powertoold

[hhobeika]

10 hours ago, powertoold said:

My 500V offboard kgsshv currently has two rev 0.4 amp boards and a rev 0.6 power board.

I am looking to swap one of the rev 0.4 amp boards to a rev 0.5 board.

When I change from a rev 0.4 to a rev 0.5 amp board, I will need to:

1) Not cross the 50K resistors

2) Not cross the 1n914 diodes

3) Not flip the O+ and O- bias outputs

Is this correct and also is there anything else I need to do when going from rev 0.4 to rev 0.5 boards?

Thanks y'all!

For rev 0.4 which is the one I built, it was either to cross the 50k resistors or the 1N914 diodes to correct the reversed feedback, not both.  That also resulted in the output polarity O+ and O- being reversed.

I am not sure if it was corrected in rev 0.5 but it should be easy enough to check by comparing the PCB traces between rev 0.4 and rev 0.5 if you have both.

[powertoold]

1 hour ago, hhobeika said:

I am not sure if it was corrected in rev 0.5 but it should be easy enough to check by comparing the PCB traces between rev 0.4 and rev 0.5 if you have both.

I see, thanks! I'm not very knowledgeable with circuits, so I'm just making sure my thoughts are correct before something blows 😅

[powertoold]

Just in case anyone needs this for future reference, I've posted the tracings of the 0.4 and 0.5 offboards. The difference is shown near the purple dot.

The front side of the boards are identical to me. Also, there's no need to cross the diodes or the 50K resistors on the 0.5. This means there's no need to flip the 0+ and O- as well.

[powertoold]

Hopefully this helps someone in the future.

My 500V offboard KGSSHV (running at ~10mA, 100R for the current adjusting resistors) developed some grainy distortion on the left channel (only with some low bass reverb frequencies). After replacing all sorts of parts, I found that the distortion was due to the IXYS transistors.

I had bought 2x IXTP01N100D from Mouser in June 2020 and 4x in Jan 2021. It turns out both of these sets were "bad" parts, symptoms below:

1) Offset voltage on left channel started at around +35V and drift upwards to around +70V

2) There was no way to lower the offset voltage, as the trimpot was maxed out

3) The sound was a bit metallic from the left channel

4) The volume was a bit lower than normal

Fortunately, spritzer sold me some of his old stock IXTP01N100D, and after replacing the new IXYSes with spritzer's old stock IXYS, all of the symptoms went away, and the amp sounded "normal" again.

With the old stock IXYS:

1) My offset voltage started around -35V and then drifted towards 0V

2) Amp seemed cooler to the touch

BTW, I think the 500V KGSSHV offboard at 10mA is the best pairing for the 007 mk1.

Edited February 10, 2021 by powertoold

[Pars]

Are Mouser or IXYS going to do anything about it, or have they been told?

[powertoold]

3 hours ago, Pars said:

Are Mouser or IXYS going to do anything about it, or have they been told?

I don't have a means to test the IXYS transistors I got from Mouser. The Jan 2021 ones I got from Mouser are lot TS1811 S23728.

If y'all got a 500V with IXYS and offset voltage problems, perhaps you have these problematic transistors as well. The amp works and all. It just doesn't sound or measure like what it's supposed to.

Edit: on closer inspection, it seems the parts I bought in June 2020 and Jan 2021 are the same lot number.

Edited February 10, 2021 by powertoold

[starcat]

I have bought from Mouser some IXYS IXCP10M90S for my T2 just some weeks ago, hope they aren't lemons as the whole PSU is full of them (13x I think). 

[jokerman777]

I was doing a little bit of layout planning on my KGSSHV build and have this newbie question about grounding.. I will be building it out of the partial kit from Mjolnir audio, the on-board heatsink version with 450V rails, other than the original PCB set I think I'm also going to make a small breakout board to distribute power from the PSU board to the two amp boards.

I have the following 2 grounding configurations in mind and wonder which one is better? Config A both the input signal ground from the XLR jacks immediately return to earth ground, then they meet again on the breakout board. Config B the input signal grounds keep separated until they met on the breakout board then return to earth ground from there. In my mind config A has the shortest ground loop between the source and the amp; Config B however is more of what star grounding is like?  Any input or suggestion much appreciated 🙂

Grounding A:

Grounding B:

[hhobeika]

On 3/23/2023 at 3:45 PM, jokerman777 said:

I was doing a little bit of layout planning on my KGSSHV build and have this newbie question about grounding.. I will be building it out of the partial kit from Mjolnir audio, the on-board heatsink version with 450V rails, other than the original PCB set I think I'm also going to make a small breakout board to distribute power from the PSU board to the two amp boards.

I have the following 2 grounding configurations in mind and wonder which one is better? Config A both the input signal ground from the XLR jacks immediately return to earth ground, then they meet again on the breakout board. Config B the input signal grounds keep separated until they met on the breakout board then return to earth ground from there. In my mind config A has the shortest ground loop between the source and the amp; Config B however is more of what star grounding is like?  Any input or suggestion much appreciated 🙂

Grounding A:

Grounding B:

In my KGSSHV build, I used something similar to your grounding B scheme, and it's very quiet with no noise or hum whatsoever.  I did not however use a series resistance and capacitance between the earth and the ground.  They are directly connected.  The XLR input connectors have a separate earth connection that I connected to the central earth connection.

[jokerman777]

I finally gotten it put together and powered it on for adjustment, high and low supply voltages are fine and all LEDs lights up.

 

I observe the following however:

-The balance adjustment I can get it drift around 0V but likely only down to the accuracy of +-0.XV at best

-The offset adjustment I can only trim the pot all the way down and get +25V and +35V on each channel, it won’t go lower and seems to be slowly drifting upwards over time.

-Bias TP measures 600V instead of 580V (I can maybe try other zeners to adjust this)

 I built it with A1968 CCS, testing is done without headphone plugged in nor is servo connected, unit was on for about 30mins when doing measurement.

 

I suppose at least this DC offset suggests I'm doing something wrong here? I cannot trim it around zero and nowhere near.

Also before any adjustment I checked that all the pots are in middle position, when I turn the amp on the first time DC offsets are as high as 150V - was that an extreme value for the offset pot?

Edit: msged Birgir and was told changing the resistor value in series with the offset pot should do it and rest is fine, sorry for dumb questions bombing 

 

[mwl168]

Nothing seems out of ordinary to me based on your description. If the offset trim pots are responding but you cannot get the offset to 0V, most likely you just need to trim the value of the resistors in series with the trim pots. 

I am going by memory without seeing the schematic but this information is in this thread somewhere.
 

[jokerman777]

🙂Got the amp finished and running fine for more than a week by now, finally got to experience some ES setup.

Many thanks for all the helpful advices I received here and there on the forum, thanks to KG for the great design and also Birgir who spent extra time clarifying a lot of my confusions in addition to supplying the partial kit.

For offset I ended up paralleling a couple of 10K resistors on the back of the board to the R in series with the trimmer, and with that I'm able to trim the offset down around 0.

For the currently back-ordered parts:

I used STP8NK80ZFP instead of FQP8N80C, and 5027 instead of 5026MOS and it has been working fine. 5027 has higher capacitance but I suppose it matters less when it's only used in the PS regulator here.